A visual brain map for AI memory systems. See your entire memory hierarchy at a glance, search semantically, edit inline, and manage priority-based loading rules.
Built on Supabase + Vercel. Works with Claude Code, Claude Desktop, ChatGPT, or any MCP-compatible AI tool.
Brain Map — Visual tree showing memory hierarchy across 3 tiers:
- Always Load (rules, recent context)
- Table of Contents (reference docs, runbooks, integrations)
- On-Demand (project, technical, personal memories by category)
Semantic Search — Find any memory by meaning, not just keywords
Inline Editing — Click any memory node to expand, edit, or delete
Editor Tab — Manage source-of-truth documents organized by load priority
Google OAuth — Secure login with email allowlist
Priority Load Order — P1 through P4 system tells AI agents what to read first
- Create a Supabase project
- Clone this repo
- Replace
YOUR_SUPABASE_URLandYOUR_SUPABASE_ANON_KEYinindex.html - Add your email to
ALLOWED_EMAILS - Deploy to Vercel
Full setup instructions in CLAUDE.md (readable by both humans and AI agents).
Not all memories are equal. Rules should load every session. Reference docs load when relevant. This system prevents AI context overload and ensures critical guardrails are never forgotten.
Long AI sessions cause context drift — the agent "forgets" rules loaded at the start. The /refresh command forces re-reading of P1 rules mid-session without starting over.
A structured boot sequence for AI agents: load rules first, then recent context, then table of contents. Ask what we're doing. Load specifics on-demand.
- Frontend: Vanilla HTML/CSS/JS (no build step, single file)
- Auth: Supabase Auth with Google OAuth
- Database: Supabase (Postgres + pgvector)
- Hosting: Vercel
- Memory Protocol: MCP (Model Context Protocol)
PRs welcome. If you build something cool on top of this, open a PR or issue.
MIT